paul-buerkner/brms

Only check required variables in post-processing methods

Open

#782 opened on Oct 31, 2019

View on GitHub
 (2 comments) (3 reactions) (0 assignees)R (1,402 stars) (220 forks)batch import
featuregood first issuepost-processing

Description

Say we are modeling a distributional parameter, like the example in the vignette:

zinb <- read.csv("https://paul-buerkner.github.io/data/fish.csv")

fit_zinb2 <- brm(bf(count ~ persons + child + camper,
                       zi ~ child), 
                 data = zinb,
                 family = zero_inflated_poisson())

and we want to get a fitted value for the distributional parameter for a particular predictor value:

fitted(fit_zinb2,
       newdata = data.frame(child = 2),
       dpar = "zi")

we get the following error:

Error: The following variables are missing in 'data': 'persons', 'camper'

even though the model for zi does not include those variables, which are only in the model for the count response.

This is a very minor issue, but it would be ideal if the code checked the variables in the newdata dataframe against the model for the dpar.

Contributor guide

Tech stack
None
Domain
data
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
under 1 hour
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
R programmingbrms package
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
50
Research direction
The issue is that when using fitted() with a distributional parameter (dpar), the function unnecessarily checks for all variables from the full model even though only a subset is needed. The fix should modify the variable checking logic in the post processing methods (likely in the fitted.brmsfit method) to only require variables that are relevant to the requested dpar. The relevant code is in the brms source, specifically in the methods for fitted(). A possible approach is to parse the model formula for the specific dpar and extract the variables needed.